Martian atmosphere loss

Why did Mars lose almost all its air while Venus kept a crushing atmosphere?

Stand on Mars and the air presses on you with about six millibars — less than a hundredth of what you feel at sea level on Earth. Stand on Venus and it presses with about ninety-two bars, the weight of water nine hundred metres down. Two neighbouring rocky planets, and their atmospheres differ by a factor of some fifteen thousand.

The explanation almost everyone has heard runs: Mars is small, so its gravity is weak, and it lost its magnetic field, so the solar wind stripped the air away. Both halves contain something true. Neither survives being pressed on.

Take the magnetic shield first. Does Venus have a global magnetic field? It does not — no internally generated dynamo field worth the name. And Venus orbits at about 0.72 astronomical units while Mars orbits at about 1.52, so since the wind thins with the square of distance, the solar wind arriving at Venus is roughly four and a half times denser. We are therefore asked to believe that an unshielded planet in the stronger wind kept ninety-two bars while an unshielded planet in the weaker wind lost nearly everything. Whatever decides this, it is not the shield. Spacecraft have measured escaping ion fluxes at both planets with closely related instruments and found them broadly comparable — nothing like the factor the atmospheres differ by. Some modelling even argues a magnetic field can increase loss by opening polar field lines that funnel ions out; that remains contested, but it is enough to unsettle the tidy story.

Now gravity. Mars's escape velocity is about 5 kilometres per second, Venus's about 10.4. Does that mean carbon dioxide evaporates off Mars? Ask what a molecule's speed actually is. At Martian temperatures a carbon dioxide molecule averages a few hundred metres per second, and thermal escape depends on the far tail of that distribution reaching 5,000 — vanishingly rare for a molecule that heavy. Carbon dioxide does not boil off Mars.

So how does it leave at all? Only by being given energy it does not have thermally. Sunlight breaks the molecule up and the recombining fragments can kick an oxygen atom out at escape speed; or the molecule is ionised, caught by the electric field the solar wind carries, and flung away. Every such channel must clear a threshold, and the threshold is exactly the escape velocity. Gravity is decisive, then, but not by holding gas down against evaporation. It sets how expensive each non-thermal kick has to be, and Venus's well is twice as deep.

Here the story stops being a single tug-of-war. How much of Mars's air actually went to space? There is a clean way to ask. Argon is chemically inert — it cannot be locked into rock, so anything missing must have escaped — and escape is mass-selective, preferring the lighter isotope. MAVEN measured the ratio of argon-38 to argon-36 in the Martian air, and from its enrichment Bruce Jakosky and colleagues concluded that roughly two-thirds of the argon Mars ever had has been lost to space. But how much of the carbon dioxide went the same way, rather than into carbonate rock, the polar caps, or the dust, is genuinely unsettled and actively argued.

Now turn the question round, because it has been asked backwards. Why does Venus have so much air? Not, chiefly, because Venus kept it. Earth holds a comparable quantity of carbon dioxide — it is simply not in the sky. It is limestone. Bake Earth's carbonate rocks and you would return tens of bars of it to the atmosphere, the same order as Venus carries. That burial needs liquid water: rain weathers silicate rock, the products carry carbon to the sea, and it settles as carbonate. Venus lost its water — the deuterium-to-hydrogen ratio in what vapour remains is more than a hundred times Earth's, the signature of hydrogen escaping while heavier deuterium lagged. With the water went the sink, and the carbon dioxide volcanoes kept exhaling had nowhere to go but the air.

Mars, meanwhile, is small in a second way that has nothing to do with escape velocity. A small body has more surface for its volume, so it sheds interior heat faster; its volcanic engine largely wound down. Loss continued, and resupply did not.

Three refinements tie the steps together. First, "stripped by the solar wind" is not wrong, it is incomplete: the wind supplies energy for several escape channels, but whether a kicked particle actually leaves is set by the planet's gravity — which is why the same wind produces such different outcomes.

Second, thickness of atmosphere is not the same quantity as retention of volatiles. Venus's ninety-two bars state where its carbon is stored, not how little it has lost — Venus lost an ocean. The two planets are not opposite results of one process; they are results of different processes compared by the same number.

Third, the honest caveats. The split of Mars's original carbon dioxide between space and rock is not settled, nor is the sequence — whether the atmosphere thinned mostly in a violent early era of intense solar activity or drained steadily over four billion years. The dynamo's shutdown around four billion years ago probably did raise the loss rate. It was a contributing cause, not the cause.

• Why hydrogen escape is the master valve for a planet's water, and what the deuterium ratio can and cannot tell us about how much Venus had.
• What the carbonate-silicate cycle does over hundreds of millions of years, and why it is called Earth's thermostat.
• Whether thickening Mars's atmosphere is physically possible at all, given how much of the inventory may no longer be on the planet.

Every term the collection defines is gathered in the glossary.